restored the Model B1 American System-Built residence at 2714 West Burnham Street in Milwaukee. In furnishing the house, the board
found no design for a living-room armchair; they decided to make use of the design found on the interior renderings of the similar (and
neighboring) Model C3 house. These designs date from 1915-1917.
As this furniture piece was apparently never constructed, and an extensive search for documentation at Taliesin uncovered no working
drawings, the chair would have to be built using only the two views which appear in the literature: an interior perspective of the C3
library/living room, and an isometric-perspective plan view of the house.
The board of Frank Lloyd Wright® Wisconsin engaged Stafford Norris III (well-known by now to the Wright community) to recreate that
chair -- in walnut -- and he in turn presented me with the challenge (and delightful opportunity) of making shop drawings for his use.
Thus, it was necessary to look closely at the two views of the chair. Two facts soon became clear: The plan view does not agree with the larger side view,
and this larger drawing lacks internal consistency. Indeed, there is evidence that the chair may have been hastily drawn; there are several instances of faulty
perspective. Nevertheless, there is enough information with which to concoct a reasonable facsimile of the architect's intention -- with some educated guess-
work and a few assumptions !
In the above detail of the isometric drawing we see that this version of the armchair (a sort of cube
chair, a form which Wright had previously explored) has four stout legs and what appear to be panels
of spindles on three sides. An earlier design, built for the Francis W Little residence, in 1903, has
somewhat similar details -- with the addition of a semi-octagonal rear "bay" not unlike the one seen
in the larger view of our C3 chair:
We'll come back to this chair later. In the meantime, two other commissions, from the same period as our ASBH house(s), show furniture details related to those
of the tables (and kitchen benches) found in the drawings above and at the B1 house, which retains its built-ins and some furniture. The same panel of square
spindles resting on a "sill" stretcher, which we see in the dining/library table and in our subject chair, are found in these pieces:
One more drawing comes into play, here -- a view (identical in perspective to the C3 chair) with remarkable similarities to its front
quarters but with a different back and projecting arms. This isn't a cube chair at all, yet it clearly belongs with the other design,
with its panel, banks of spindles, and six slab legs. This drawing has been given a modern Taliesin catalog number in black ink --
0519.001, dating to 1905 -- fully a decade earlier than the ASBH work. Can this be right ?
Now let's look closely at our main drawing. We see that the chair appears to equal the table in height -- a useful datum. We see that the top panel has a semi-circular
cut-out for the seat back, so we feel comfortable referring to the isometric plan view, which suggests that this is intended to be a true arc, in accordance with Wright's
custom (he never ever designed with ellipses, preferring shapes his compass and other essential tools could produce). The aberrant curved shape shown in the drawing
below should be a first clue that something's not right, here: note that a projection of the rear legs places them in an impossible location at the top of the chair.
(There's nothing wrong with that beautiful foliage, though, is there !)
Further study reveals inconsistencies in perspective; several horizontals in the chair -- and the table -- deviate from true (at B, C, and D, below), leaving us in doubt as to the sincerity of the draftsman, and complicating our job of interpretation. However, this condition has an upside: it gives us the liberty to take the drawing with a "grain of salt," permitting the interpreter (by necessity) to choose among the further conundrums ahead. To wit: How do we reconcile the lack of projection of the arm furthest from the viewer, with the clear projection of the one nearest ? What do we make of the continuous vertical edges of the middle leg, which the front one lacks as it meets the bottom rail ? Could that down-sloping rail have been drawn thus as a last-minute save from an error in placement of the termination of both legs at the floor -- when the drafter intended that front leg to end up inboard of the middle one, and thus shorted (as it is) -- in perspective ?
I chose to read it thus. My take on this chair is that, below the arm plate, the designer looked for ways to incorporate the cylinder implied by the semi-circular seat back into
the assemblage of legs and spindles -- leaving only the framed panels at the front in a forthrightly orthogonal condition. The fact of the three sets of legs spreading in the middle and tucking in, fore and aft, furthers this subtle cylindricality.
The last question, following from the above, is whether the bottom of the semi-octagonal spindled bay is intended to slope
upward, as it appears to do in the perspective. In the annotated drawing above I completed the construction of this part of
the chair, proving (to myself) that this was the intention. Only one straw remained to save me (and the constructor) from this
very inconvenient reading: a single vertical line (at A) which, I believe, is intended to be the upper edge of the right-hand angled
return of that bottom plate. I clutched at that straw -- with the result you see in my measured drawings below.
1In the plan, the lower portion is a section taken below the arm plate while the other represents a section through the chair
below the seat. Plywood is 1/2" while most solid stock is 7/8".
2
3The bottom plate is two 1/2" walnut panels laminated; the bottom rail of the panel section is a matching 1" piece. The
1/2" kerf-bent plywood seat-back shell engages a 3/4" seat panel.
4This full-size detail shows that the design arranged itself simply on a grid of 1/2" by 7/8":
5The plan drawing shows two different configurations to the rear-most spindles; in one version there is no spindle at the 45º return, while the other is provided
with one. The Little chair (detail below) gives us a precedent, supporting the second version. Below is the full-size detail sheet which provides both alternatives.
The spindles and their spaces are 7/8" throughout in the first (darker) option, while in the second (lighter) the spindles remain at 7/8" square while the spaces
shrink to 5/8" and 9/16" respectively (to maintain the same chair width). Alternate centerlines provide for increasing the chair width if desired.
6
This sheet shows three options for the configuration of the arm plate, each one made from planks of 7" width. The plywood shells have different grain
orientation relating to their construction. The loose upholstery shell could be installed with Velcro tabs or strips, as shown in the full-size detail (sheet 5).
7SDR